Covid Hasn’t Disrupted Educational Progress. Test Scores Were Already Falling.

No, Covid hasn’t “erased two decades” of progress in education. Dig deeper into national test scores and you’ll find there’s been little or no progress for most of that time.

Last week, officials released test scores in reading and math for a representative sample of nine-year-olds across the country and declared that, compared to nine-year-olds tested just before the pandemic, their performance had declined sharply. The press release noted that this was “the largest score decline in reading since the 1990s and the first-ever decline in mathematics.”

The news led to dramatic media coverage. “The Pandemic Erased Two Decades of Progress in Math and Reading,” according to the headline in The New York Times. “Test scores in elementary school math and reading plummeted to levels unseen for decades,” the Washington Post reported. Education news outlets used similar language.

These reports suggest that the recent history of American education was a sunny tale of continuous improvement until Covid derailed it. Alas, as officials overseeing the national tests themselves recognize, test scores have been stagnant or declining since at least 2009, and gaps between high- and low-scorers have been widening.

To be sure, the recent scores are alarming. But it’s important to understand that students weren’t making progress before the pandemic, so that we have a better understanding of how to undo the damage.

A little background: The recently released scores came from a battery of tests called the National Assessment of Educational Progress, or NAEP. They fall into two basic categories. First, there are the “Main NAEP” tests, which have been around since the early 1990s. They’re given every two years to fourth and eighth graders in reading and math (and sometimes to twelfth graders), and intermittently in other subjects. (A planned administration in 2021 was delayed due to the pandemic and is being given this year.)

The second category is the Long-Term Trend NAEP, or LTT, which has been given periodically since the early 1970s. Rather than being administered to students at certain grade levels, the LTT is given to students at ages nine, thirteen, and seventeen. The questions on the LTT don’t change as much as those on the Main NAEP, to enable more accurate comparisons across time.

The recently released scores are for the LTT, and the only students who took it were nine-year-olds. That age group had been tested in January 2020, shortly before the pandemic hit, and those who oversee the NAEP decided to test nine-year-olds again in January 2022 to shed light on the effects of school closures and remote classes.

This year’s LTT did show a steep decline in performance and growing gaps between high- and low-scorers since 2020. But if you go back to 2012, when the LTT was given to all three age groups, there already wasn’t much progress to see—at least, not recent progress. Scores for nine and 13-year-olds in 2012 had increased since the early 1970s, but only 13-year-olds had made gains since the previous LTT in 2008. And 17-year-olds had made no progress since the early 1970s.

Similarly, there had been progress in narrowing gaps between racial groups in earlier years, but it had stalled. There was no change since 2008 in either the Black/white or Hispanic/white gaps, except for some narrowing for Hispanic 13-year-olds in reading.

Scores from the Main NAEP tell much the same story. As two officials connected with NAEP wrote in Education Next, characterizing the recent score drop as being just about Covid “ignores the complete picture of student achievement in the decade leading up to the pandemic.”

In math, there was significant progress on Main NAEP between 1990 and 2009. In reading, however, there was only modest progress until 1998, followed by two decades of stagnation. On the most recent Main NAEP in 2019, math scores were mixed, and reading scores actually declined.

And in recent years, the scores of lower-scoring students have gotten worse, at all grade levels, while higher-scoring students have either improved or held steady. The declines in both reading and math in 2019 made the scores comparable to those in 2005, according to the officials writing in Education Next, and the performance of low-scorers in grade 12 reading was the worst since the test was first given in 1992.

In other words, we’ve seen this movie before. The 2022 results are different only in that scores have now declined further, and they’ve dropped even for high-scoring students. The gap between high- and low-scorers has continued to grow, however, because low-scorers dropped more sharply.

It’s important to see the recent scores as just one more downward step in a decline that’s been going on since at least 2009. If you see them as a sharp break from a previous record of steady progress, as news reports have suggested, you’re liable to think the solution is for schools to double down on what they were doing before the pandemic. Indeed, that’s what most experts are recommending: more school time, more tutoring, more summer and after-school programs.

But if you realize that what we were doing before wasn’t actually working, at least for low-scoring students, that should lead you to believe we need to change something more fundamental—like what we are teaching and how.

The basic problem in American education is that, for complex reasons, it has been cut off from what scientists have discovered in recent decades about how learning actually works. The result is that, often, the only kids who thrive in school are those who would probably learn no matter what, generally because they have more highly educated parents (and it appears even many of them have struggled during the pandemic). The others are ill-served by our system and never get a chance to realize their true potential.

I don’t have space to detail everything that would need to change to bring education in line with cognitive science, so let’s focus on reading. One problem there is that, because of deficiencies in teacher training, many children never learn to decipher or “decode” words—at least, not well enough to do it automatically, which enables fluent reading. Evidence indicates that most low-scoring younger students haven’t learned to decode fluently. That’s presumably true for many at higher grade levels as well, but because NAEP purports to test comprehension rather than decoding, we don’t have good data on their numbers.

But even kids who can decode often lack the academic knowledge and vocabulary needed to understand complex written text. They should be acquiring those things beginning in elementary school, but the standard approach to reading comprehension prioritizes supposed skills like “finding the main idea,” which students practice using easy-to-read texts on random topics. Elementary and sometimes middle schools devote little time to content-rich subjects like history and science, leaving students ill-equipped to understand the texts they’re expected to read in high school.

All of that largely explains the stagnation and decline in reading scores, especially in high school, and the growing gap between low- and high-achievers. But there was some progress in math until 2009, and reading scores have been particularly depressing since then. What happened?

It’s impossible to know for sure, but one widespread change around that time was the adoption by states of more rigorous academic standards in reading and math. The idea was to raise academic achievement by raising expectations. But that’s not what happened.

High expectations are crucial, but if you don’t give students the information and explicit teaching they need to meet them, they can backfire. Material can go over students’ heads—or teachers, seeing that kids aren’t able to understand what they’re supposed to be learning, can water down the material. I suspect some combination of those things has happened.

The solution, from what I’ve seen, is for schools and districts to adopt content-rich, logically structured, knowledge-building curricula that begin in the early elementary grades—and provide teachers with effective support to implement them. There are perhaps half a dozen such curricula available now, and their use is growing, although it’s still far from the norm.

Teachers who have switched to this fundamentally different approach are often amazed by what their students can do. And while the results may not be apparent in test scores for quite some time, this radical break with past practice is the only reliable way to boost those scores—and to help millions of kids recover from the damage Covid has wreaked on an education system that was already fundamentally flawed.